Fauziah Roshafara
Prodi Statistika, Fakultas Farmasi dan Sains, Universitas Islam Bandung, Indonesia

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

Klasifikasi Sentimen Publik Terhadap Banjir di Sumatera pada Teks Berita dan Media Sosial X Menggunakan IndoBERT Marini Salmonia Kisa; Fauziah Roshafara
Bandung Conference Series: Statistics 89-98
Publisher : UNISBA Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/bcss.v6i2.24414

Abstract

Abstract. Flood disasters on Sumatra Island have triggered diverse public responses through online news coverage and social media platform X. This study applies IndoBERT, a BERT-based model that uses pre-trained weights and is subsequently fine-tuned to classify public sentiment regarding the disaster. Data were collected through web scraping from three national news portals, namely Detik, CNBC Indonesia, and Kompas, as well as from social media X using the keyword “banjir Sumatera” between November 28, 2025, and January 9, 2026. After data selection, 2,687 records were obtained, consisting of 1,494 news texts and 1,193 X posts. This study aims to evaluate IndoBERT performance on two data types with different linguistic characteristics. Initial sentiment labeling was conducted using the InSet lexicon, producing three classes: negative, neutral, and positive. On news data, IndoBERT achieved an accuracy of 0.81, precision of 0.81, recall of 0.80, and F1-score of 0.80. On social media X data, it achieved an accuracy of 0.79, precision of 0.78, recall of 0.78, and F1-score of 0.78. These findings indicate that IndoBERT performs better on news texts because their formal language characteristics are more consistent with the model’s pre-training corpus. The results demonstrate that corpus compatibility influences model performance across different Indonesian textual domains. Abstrak. Bencana banjir di Pulau Sumatera memunculkan beragam respons publik melalui pemberitaan media berita online dan media sosial X. Penelitian ini menerapkan IndoBERT berbasis arsitektur BERT dengan memanfaatkan bobot pre-trained yang kemudian digunakan pada proses fine-tuning untuk mengklasifikasikan sentimen publik terkait bencana tersebut. Data dikumpulkan melalui web scraping dari tiga portal berita nasional, yaitu Detik, CNBC Indonesia, dan Kompas, serta media sosial X menggunakan kata kunci “banjir Sumatera” pada periode 28 November 2025 hingga 9 Januari 2026. Setelah proses seleksi, diperoleh 2.687 data yang terdiri atas 1.494 teks berita dan 1.193 cuitan X. Penelitian ini bertujuan mengevaluasi kinerja IndoBERT pada dua jenis data dengan karakteristik yang berbeda. Pelabelan awal dilakukan menggunakan kamus leksikon InSet dan menghasilkan tiga kelas sentimen, yaitu negatif, netral, dan positif. Hasil evaluasi menunjukkan bahwa pada data berita IndoBERT memperoleh accuracy 0,81, precision 0,81, recall 0,80, dan F1-score 0,80. Sementara itu, pada data media sosial X diperoleh accuracy 0,79, precision 0,78, recall 0,78, dan F1-score 0,78. Hasil tersebut menunjukkan bahwa IndoBERT lebih baik dalam mengklasifikasikan sentimen pada teks berita karena karakteristik bahasanya lebih formal sesuai dengan korpus pre-training IndoBERT.
Pemodelan Kemiskinan di Provinsi Jawa Barat dengan Geographically Weighted Regression (GWR) Kurnia Ardi Ferdianto; Fauziah Roshafara
Bandung Conference Series: Statistics 243-250
Publisher : UNISBA Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/bcss.v6i2.25414

Abstract

Abstract. Poverty in West Java Province in March 2024 was recorded at 7.46 percent, or 3.89 million people, with patterns that vary across regions due to spatial effects. This condition indicates that classical regression, which assumes constant parameters across all regions, is less appropriate to use, making it necessary to apply Geographically Weighted Regression (GWR), a method capable of accommodating local variations in relationships between regions. This study aims to model the number of poor population across 27 regencies/cities in West Java Province in 2024 using GWR, with independent variables including the Gini ratio, GRDP growth rate, per capita expenditure, the Community Literacy Development Index (IPLM), minimum wage, and the Labor Force Participation Rate (TPAK). The spatial weighting used an adaptive bisquare kernel function, with the optimum bandwidth determined through the Cross Validation method, and data processing was carried out using R Studio. The model fit test results show an F-value (2.3704) greater than the F-table value (2.235) with a p-value of 0.03764, indicating that GWR is more appropriate than OLS. The GWR model produced different equations for each region, with an AICc value of 33.7708, and significant variables that varied across regions, grouped into six clusters of regencies/cities. Abstrak. Kemiskinan di Provinsi Jawa Barat pada Maret 2024 tercatat sebesar 7,46 persen atau 3,89 juta jiwa, dengan pola yang bervariasi antarwilayah akibat adanya pengaruh spasial. Kondisi ini menunjukkan bahwa regresi klasik yang mengasumsikan parameter konstan di seluruh wilayah kurang tepat digunakan, sehingga diperlukan metode Geographically Weighted Regression (GWR) yang mampu mengakomodasi variasi hubungan antarwilayah secara lokal. Penelitian ini bertujuan memodelkan jumlah penduduk miskin di 27 kabupaten/kota Provinsi Jawa Barat tahun 2024 menggunakan GWR, dengan variabel independen meliputi rasio gini, laju pertumbuhan PDRB, pengeluaran per kapita, IPLM, upah minimum, dan TPAK. Pembobot spasial menggunakan kernel adaptive bisquare dengan bandwidth optimum ditentukan melalui metode Cross Validation, dan pengolahan data dilakukan menggunakan R Studio. Hasil uji kesesuaian model menunjukkan F-hitung (2,3704) > F-tabel (2,235) dengan p-value 0,03764, sehingga GWR lebih sesuai digunakan dibandingkan OLS. Model GWR menghasilkan persamaan berbeda untuk tiap wilayah dengan nilai AICc sebesar 33,7708, serta variabel signifikan yang bervariasi antarwilayah, dikelompokkan menjadi enam kelompok kabupaten/kota.
Peramalan Harga Emas Menggunakan Geometric Brownian Motion Rifa Fadhila; Fauziah Roshafara
Bandung Conference Series: Statistics 291-300
Publisher : UNISBA Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/bcss.v6i2.25726

Abstract

Abstract. Gold is one of the most popular investment instruments because its price fluctuates due to various economic factors. Therefore, a method capable of modeling stochastic price movements is needed. Geometric Brownian Motion (GBM) has been widely used to model financial asset prices because it incorporates both deterministic and stochastic components. This study aims to determine the accuracy of the GBM model in forecasting gold prices based on the Mean Absolute Percentage Error (MAPE) and to forecast gold prices for the next five periods. Daily gold price data from 1 January 2024 to 31 December 2025, obtained from Investing.com, were used in this study. The data were divided into 415 training observations and 104 testing observations. Model parameters were calculated based on return values, and Monte Carlo simulations were performed using 100, 500, and 1000 iterations. Model accuracy was evaluated using MAPE, and the model with the smallest MAPE value was selected for forecasting. The results showed that the Monte Carlo simulation with 500 iterations produced the best model with a MAPE value of 7.8838%. Based on this model, gold prices were projected to increase from 4,321.53 USD/oz in the first period to 4,344.99 USD/oz in the fifth period. These findings indicate that the GBM model provides good forecasting accuracy and is a suitable alternative for short-term gold price forecasting. Abstrak. Emas merupakan salah satu instrumen investasi yang nilainya berfluktuasi akibat berbagai faktor ekonomi, sehingga diperlukan metode yang mampu memodelkan pergerakan harga emas secara stokastik. Penelitian ini bertujuan untuk mengetahui tingkat akurasi model Geometric Brownian Motion (GBM) dalam melakukan peramalan harga emas berdasarkan nilai Mean Absolute Percentage Error (MAPE) serta memperoleh hasil peramalan harga emas untuk lima periode ke depan. Data yang digunakan berupa harga emas harian periode 1 Januari 2024–31 Desember 2025 yang diperoleh dari Investing.com. Data dibagi menjadi 415 data training dan 104 data testing. Parameter model dihitung berdasarkan nilai return, kemudian dilakukan simulasi Monte Carlo sebanyak 100, 500, dan 1000 iterasi. Akurasi model dievaluasi menggunakan MAPE, kemudian model terbaik digunakan untuk melakukan peramalan. Hasil penelitian menunjukkan bahwa simulasi Monte Carlo dengan 500 iterasi menghasilkan model terbaik dengan nilai MAPE sebesar 7,8838%. Berdasarkan model tersebut, harga emas diproyeksikan mengalami tren meningkat selama lima periode ke depan, yaitu dari 4.321,53 USD/oz pada periode pertama menjadi 4.344,99 USD/oz pada periode kelima. Hasil penelitian menunjukkan bahwa model GBM mampu memberikan tingkat akurasi yang baik untuk peramalan harga emas jangka pendek.